An observational approach to tunnel design and construction is commonly employed in order to assess excavation driven displacements and to verify the design of temporary support systems utilized to control and minimize surface deformation. The umbrella arch is such a support system, composed of longitudinal support members, which provide stability to both the working face and region ahead of the excavation. However, a distinct lack of knowledge exists in terms of the ground-support interaction and performance beyond the working face. Within this context an application of a novel distributed optical strain sensing technique in combination with umbrella arch support members is presented as a supplemental tool to conventional monitoring techniques of the observational approach with the capability of "seeing" past the working face.
During the design and construction of tunnels in weak ground, a temporary support regime is often considered which is utilized in order to provide provisional support until the final liner is installed. Such support systems are primarily designed contingent upon the anticipated ground conditions as well as the project specified requirements and limitations. Considering the example of shallow urban tunnelling, the control and mitigation of surface settlements will be of primary concern. The assortment and arrangement of potential support systems installed in order to abide by such project demands involve but are not limited to: umbrella arch support (i.e. forepoles, spiles), ground freezing, face and radial bolting, steel sets, shotcrete, inverts, etc. (Figure 1). An observational tunnelling method, as provided by the Austrian Society for Geomechanics (2010), provides a design rational that integrates the surrounding rock into the overall support structure (i.e. the supporting formations will themselves be a part of the supporting structure as the rock is able to support itself to a certain degree) (Romero 2002). By permitting a controlled deformation of the ground mass (i.e. a limited strain of approximately 1%) stresses are provided with an opportunity to be partly released, becoming less stiff, and ultimately allowing a less cost (and time) intensive support system to be implemented (Kontogianni & Stiros 2002). However, this deformation based, observational tunnelling approach requires that the mechanisms of the ground and ground-support interaction be well understood in terms of the three-dimensional (3D) tunnel effects / considerations. Furthermore, a monitoring program capable of capturing such behaviour must be implemented as to verify or falsify the assumptions made during the design stage (Schubert 2008).